The off-season maize crop (named as "Safrinha") has been frequently suffering significant yield losses due to Phaeosphaeria leaf spot in state of São Paulo, Brazil. As a strategy for minimizing these losses, agrometeorological disease forecasting models can help in the planning and decision-making for disease control. Therefore, this work has as objective to develop and validate agrometeorological models for forecasting the percentage of the leaf area affected by the disease (% AFA) in function of the daily accumulation of average temperature and rainfall, in a regional scale. For that, information of Phaeosphaeria leaf spot severity in 136 maize cultivars with different resistance levels, during the "safrinha", in 14 different places of the State of São Paulo were used. Models were developed for susceptible, moderately resistant and resistant cultivars and the results showed good performance with high values of determination coefficient (0.92, 0.81 and 0.83 respectively) and index of agreement of Willmott (0.98, 0.87, 0.94 respectively). Besides, test F was significant at 1% of probability for all cases. The results also allowed to suggest that rainfall and air temperature are the preponderant factors controlling Phaeosphaeria leaf spot occurance in the State of São Paulo, Brazil. These models present the advantage of using parameters usually obtained in weather stations and they can be used in early disease warning systems in the State.
disease; Phaeosphaeria maydis; Zea mays; climate; rainfall; temperature
